Run 182: findings from the verifying run after the run 181 batch #425

Closed
opened 2026-11-08 03:12:38 -05:00 by cmoriarty · 12 comments
Owner

Run 182 is the verifying run after the run 181 batch (PR #421, #423), the Scratch brief on cmoriarty/scratch-vllm-421, thorough, judge at medium (braid#420), every other step at xhigh. Baselines: run 174 (judge at medium, 225 agent min) and run 181 (judge at xhigh, 362 agent min). Findings land here as comments while it runs; the local log is docs/design/measurements/work/run-182/findings.md.

Scorecard for the batch (filled in as the steps land):

  • #412 steer carries on / apply continues its session: n/a unless steered
  • #413 apply reads screenshots at ../screenshots: ?
  • #416 judge runs the pasted motion probe: ?
  • #418 code review's reads of pasted files (181: 12 in 28 reads): ?
  • #414 round 2 honours the look list: ?
  • #415 clock holds socket state, filmstrip possible: ?
  • #417 gradient false positive gone: ?
  • #419 browser check length (181: 5–7 min, 80 file reads): ?
  • #420 judge requests carry medium: ?

So far (planning, 16 min)

  1. Explore 1.6 min, propose 4.6 min (181: 8.3) with a quarter of the reasoning tokens, combined review 8.6 min (181: 7.9), all at xhigh. Explore chose Socket.IO over Colyseus (181's decorators cost its apply an hour).
  2. The review's scope lens noted the window.__app hook as "not asked for by the brief". The brief's rule asks the proposal for it (#407); the lens wrote a note, the revise step left it, so the hook survived. A should-fix on the same ground would strip the judge's way in: the review brief should name the hook rule as Braid's.
  3. Revise fixed all six should-fix in 2 min: five missing Depends on: lines (the propose brief could ask up front) and a random-first-dealer scenario that passed without randomness.
  4. Ask's first use was refused by the delivery (route 202, delivery "Repair is for a stopped or failed run"): fixed in PR #423. Then Ask's answer showed in Repair's panel: #424.
  5. The run was stopped at 03:03 during openspec.apply[all] attempt 1 (about 70 min, 35+ turns) to deploy #423, and resumed at 03:04 on attempt 2. Apply time and attempts are not comparable with run 181's; the judge rounds and the review are.
Run 182 is the verifying run after the run 181 batch (PR #421, #423), the Scratch brief on `cmoriarty/scratch-vllm-421`, thorough, judge at `medium` (braid#420), every other step at `xhigh`. Baselines: run 174 (judge at medium, 225 agent min) and run 181 (judge at xhigh, 362 agent min). Findings land here as comments while it runs; the local log is `docs/design/measurements/work/run-182/findings.md`. **Scorecard for the batch** (filled in as the steps land): - #412 steer carries on / apply continues its session: n/a unless steered - #413 apply reads screenshots at `../screenshots`: ? - #416 judge runs the pasted motion probe: ? - #418 code review's reads of pasted files (181: 12 in 28 reads): ? - #414 round 2 honours the look list: ? - #415 clock holds socket state, filmstrip possible: ? - #417 gradient false positive gone: ? - #419 browser check length (181: 5–7 min, 80 file reads): ? - #420 judge requests carry `medium`: ? **So far (planning, 16 min)** 1. Explore 1.6 min, propose 4.6 min (181: 8.3) with a quarter of the reasoning tokens, combined review 8.6 min (181: 7.9), all at xhigh. Explore chose Socket.IO over Colyseus (181's decorators cost its apply an hour). 2. The review's scope lens noted the `window.__app` hook as "not asked for by the brief". The brief's rule asks the proposal for it (#407); the lens wrote a note, the revise step left it, so the hook survived. A should-fix on the same ground would strip the judge's way in: the review brief should name the hook rule as Braid's. 3. Revise fixed all six should-fix in 2 min: five missing `Depends on:` lines (the propose brief could ask up front) and a random-first-dealer scenario that passed without randomness. 4. Ask's first use was refused by the delivery (route 202, delivery "Repair is for a stopped or failed run"): fixed in PR #423. Then Ask's answer showed in Repair's panel: #424. 5. The run was stopped at 03:03 during `openspec.apply[all]` attempt 1 (about 70 min, 35+ turns) to deploy #423, and resumed at 03:04 on attempt 2. Apply time and attempts are not comparable with run 181's; the judge rounds and the review are.
Author
Owner

Apply, attempt 2 at 55 min (after the resume): 163 turns, context 202k with one compaction, 115k reasoning tokens. Attempt 1 had done 100 turns in 45 min before the stop.

  1. #413 holds. Every screenshot read went to /var/lib/openspecflow/runs/182/screenshots/<name> on the first try (landing, table-bidding, table-waiting-play); no read under the worktree, no search.
  2. The apply drives the browser itself, and goes through whole games. No browser subagent was started; the primary made 111 browser_evaluate, 24 screenshots, 24 tabs calls and 15 navigates, and its last words are "Driver re-launched with the fixed parser. Waiting for the game to complete." The brief says one look at what changed, not a test campaign, and to delegate to browser by default (braid#383); run 172's lesson was that the primary's looks fill its own context. Candidate: the apply brief says that a whole session by script is the judge's, and that the apply's own look is a screenshot per changed state, through browser.
  3. Two guards fired as meant. The fresh-browser rule refused the first Playwright call once (then close), and one page script was cut at 180 s with "the script did not settle; check its loop ends" (braid#329) instead of holding the step.
  4. Edit misses: 9 tool errors, most "Could not find oldString" on game.ts and the rooms tests; each cost a turn, none stalled it.

Steps 19 of 45; apply still running.

**Apply, attempt 2 at 55 min (after the resume): 163 turns, context 202k with one compaction, 115k reasoning tokens.** Attempt 1 had done 100 turns in 45 min before the stop. 6. **#413 holds.** Every screenshot read went to `/var/lib/openspecflow/runs/182/screenshots/<name>` on the first try (landing, table-bidding, table-waiting-play); no read under the worktree, no search. 7. **The apply drives the browser itself, and goes through whole games.** No `browser` subagent was started; the primary made 111 `browser_evaluate`, 24 screenshots, 24 `tabs` calls and 15 navigates, and its last words are "Driver re-launched with the fixed parser. Waiting for the game to complete." The brief says one look at what changed, not a test campaign, and to delegate to `browser` by default (braid#383); run 172's lesson was that the primary's looks fill its own context. Candidate: the apply brief says that a whole session by script is the judge's, and that the apply's own look is a screenshot per changed state, through `browser`. 8. **Two guards fired as meant.** The fresh-browser rule refused the first Playwright call once (then `close`), and one page script was cut at 180 s with "the script did not settle; check its loop ends" (braid#329) instead of holding the step. 9. **Edit misses:** 9 tool errors, most "Could not find oldString" on `game.ts` and the rooms tests; each cost a turn, none stalled it. Steps 19 of 45; apply still running.
Author
Owner

Apply done: 115 agent minutes over the two attempts (45 before the stop, 70 after), 356 turns, one compaction at 202k with 16 tasks left. Run 181's apply took 233 min over two attempts (the steer ended its first). All 38 tasks ticked, npm test 62/62, typecheck clean, ban check clean on 15 test files, 28 scenarios left to the Visual check in visual.json, 5 untested with reasons. The agent's own in-browser verification covered the table, the scoreboard, the end screen, help, a 390×844 portrait game played by taps, rematch, mute, keyboard play and reduced motion: thorough, and the reason for finding 7 (the primary drove all of it itself, through whole games).

Code review running next: #418's check (reads of pasted files, 181: 12 in 28 reads) comes with it.

**Apply done: 115 agent minutes over the two attempts (45 before the stop, 70 after), 356 turns, one compaction at 202k with 16 tasks left.** Run 181's apply took 233 min over two attempts (the steer ended its first). All 38 tasks ticked, `npm test` 62/62, typecheck clean, ban check clean on 15 test files, 28 scenarios left to the Visual check in `visual.json`, 5 untested with reasons. The agent's own in-browser verification covered the table, the scoreboard, the end screen, help, a 390×844 portrait game played by taps, rematch, mute, keyboard play and reduced motion: thorough, and the reason for finding 7 (the primary drove all of it itself, through whole games). Code review running next: #418's check (reads of pasted files, 181: 12 in 28 reads) comes with it.
Author
Owner

Code review: 6 min, 19 turns, context 52k → 117k (181: 13.9 min, 36 turns, 165k). Three small server fixes (a path-containment check, a dead re-export, open-games listing rooms whose players all left); 15 server tests and the typecheck green after them.

  1. #418 did not bind. The brief pasted 54 new files whole (146,583 characters) and named 25 more as too long to paste. The review still read 24 distinct files in 30 reads, every one of them a pasted file, and not one read was answered by the guard. Either the record under <run>/tmp/pasted.json was not written, or the guard's match (path, mtime, size) fails on production. Inspecting the run's folder next; a fix follows.

  2. The judge's brief is as the batch meant it (53,922 characters): the motion rule ("never write a freeze of your own") sits before the probes block, which sits before "Write the report as you go" and "How to end" (#416); the browser paragraph says it cannot read the source and asks for the hook's names (#419); the screenshots line names ../screenshots (#413). Its requests carry medium (#420 holds).

**Code review: 6 min, 19 turns, context 52k → 117k (181: 13.9 min, 36 turns, 165k).** Three small server fixes (a path-containment check, a dead re-export, open-games listing rooms whose players all left); 15 server tests and the typecheck green after them. 11. **#418 did not bind.** The brief pasted 54 new files whole (146,583 characters) and named 25 more as too long to paste. The review still read 24 distinct files in 30 reads, every one of them a pasted file, and not one read was answered by the guard. Either the record under `<run>/tmp/pasted.json` was not written, or the guard's match (path, mtime, size) fails on production. Inspecting the run's folder next; a fix follows. 12. **The judge's brief is as the batch meant it** (53,922 characters): the motion rule ("never write a freeze of your own") sits before the probes block, which sits before "Write the report as you go" and "How to end" (#416); the browser paragraph says it cannot read the source and asks for the hook's names (#419); the screenshots line names `../screenshots` (#413). Its requests carry `medium` (#420 holds).
Author
Owner

Correction to finding 11: #418 held. Of the review's 24 distinct reads, 0 were of the 29 files the brief pasted whole; all 24 were among the 25 files the brief listed as "not pasted: read it". The guard had nothing to answer (the record on disk holds the 29, with matching mtimes). What the finding really is:

  1. The paste budget spends itself on the wrong files. _new_file_blocks pastes in git status order until 80,000 characters, so .dockerignore, Dockerfile, README.md, package.json and every client component were pasted whole, and the engine (game.ts 427 lines, deck.ts 74, rounds.ts 13, types.ts 118), the server (rooms.ts 393, index.ts 45, static.ts 61) and all their tests were left to read, rounds.ts and the tsconfigs included though they are tiny. The review then read exactly those. Candidate: paste the source before configuration and documentation, always paste a file under a few hundred lines, and leave only package-lock.json-sized files to a read. Filed as the next batch's issue.
**Correction to finding 11: #418 held.** Of the review's 24 distinct reads, 0 were of the 29 files the brief pasted whole; all 24 were among the 25 files the brief listed as "not pasted: read it". The guard had nothing to answer (the record on disk holds the 29, with matching mtimes). What the finding really is: 13. **The paste budget spends itself on the wrong files.** `_new_file_blocks` pastes in `git status` order until 80,000 characters, so `.dockerignore`, `Dockerfile`, `README.md`, `package.json` and every client component were pasted whole, and the engine (`game.ts` 427 lines, `deck.ts` 74, `rounds.ts` 13, `types.ts` 118), the server (`rooms.ts` 393, `index.ts` 45, `static.ts` 61) and all their tests were left to read, `rounds.ts` and the tsconfigs included though they are tiny. The review then read exactly those. Candidate: paste the source before configuration and documentation, always paste a file under a few hundred lines, and leave only `package-lock.json`-sized files to a read. Filed as the next batch's issue.
Author
Owner

Judge round 1: attempt 1 lost to a stall the batch itself made possible; attempt 2 running (137 min on the step so far).

  1. A page script that pauses the clock has no time limit. At 04:28 the judge wrote its own browser_evaluate: clock.pause(); const ack = await … (a socket round trip). The clock's new gate (#415) holds that message while the clock is paused, so the promise never settled; and the 180 s bound on page scripts (braid#329) is a setTimeout in the page, which now runs on Braid's clock, paused. The call sat 1,201 s until the watchdog ended the attempt as "silent for 1198 s while busy" and redid the step: 27 minutes and the judge's first look gone. Fix in hand: the clock exposes the machine's timer (realTimeout) and the bound uses it, so the script is cut at 180 s whatever the clock does. Proved by a journey; PR next.

  2. The first browser check spent 60 turns waiting. Lobby check: landing, create, join made 26 wait_for calls (22 of them 30 s) and 28 finds for a players list that never grew (its brief promised bots the judge had not started yet), took 2 screenshots, and ran to its 60-turn cap in 13.9 min; the continuation did the whole check in 1.6 min. No file reads at all (#419's change holds), and its requests carry medium. So a check's length can be waits as well as reads: a candidate rule is a cap on wait_for per check, or a brief that forbids waiting for what the caller has not started.

**Judge round 1: attempt 1 lost to a stall the batch itself made possible; attempt 2 running (137 min on the step so far).** 14. **A page script that pauses the clock has no time limit.** At 04:28 the judge wrote its own `browser_evaluate`: `clock.pause(); const ack = await …` (a socket round trip). The clock's new gate (#415) holds that message while the clock is paused, so the promise never settled; and the 180 s bound on page scripts (braid#329) is a `setTimeout` *in the page*, which now runs on Braid's clock, paused. The call sat 1,201 s until the watchdog ended the attempt as "silent for 1198 s while busy" and redid the step: 27 minutes and the judge's first look gone. Fix in hand: the clock exposes the machine's timer (`realTimeout`) and the bound uses it, so the script is cut at 180 s whatever the clock does. Proved by a journey; PR next. 15. **The first `browser` check spent 60 turns waiting.** `Lobby check: landing, create, join` made 26 `wait_for` calls (22 of them 30 s) and 28 `find`s for a players list that never grew (its brief promised bots the judge had not started yet), took 2 screenshots, and ran to its 60-turn cap in 13.9 min; the continuation did the whole check in 1.6 min. No file reads at all (#419's change holds), and its requests carry `medium`. So a check's length can be waits as well as reads: a candidate rule is a cap on `wait_for` per check, or a brief that forbids waiting for what the caller has not started.
Author
Owner
  1. The production host rebooted under the judge, twice over. osfd was told to exit by SIGTERM at 05:41 EST (the host's update run stopping services) and the host rebooted at 05:49 onto a new kernel (5.14.0-687.61 → 687.66; the previous Sunday, November 1, rebooted at 05:27 the same way). No deploy was involved: production stayed on 2273743 and Portainer's last stack update was 03:04. Braid's restart recovery did what it should: attempt 2 (56 min of judging, four browser checks done) was adjudicated orphaned at 05:44 and attempt 3 continued it; the reboot cut attempt 3 at 05:54, and attempt 4 continued at 05:55 from Braid's note. Cost: about 15 minutes and the judge's working context twice. Not a Braid finding, but a run started on a Saturday night meets the Sunday update window; worth scheduling runs or the window around each other.

Judge checks so far (attempt 2, all at medium, none reading a file): lobby 13.9 min (the waits, finding 15) + continuation 1.6 min, table 6.2 min, scoreboard/help/mute 3.0 min.

16. **The production host rebooted under the judge, twice over.** osfd was told to exit by SIGTERM at 05:41 EST (the host's update run stopping services) and the host rebooted at 05:49 onto a new kernel (5.14.0-687.61 → 687.66; the previous Sunday, November 1, rebooted at 05:27 the same way). No deploy was involved: production stayed on 2273743 and Portainer's last stack update was 03:04. Braid's restart recovery did what it should: attempt 2 (56 min of judging, four `browser` checks done) was adjudicated orphaned at 05:44 and attempt 3 continued it; the reboot cut attempt 3 at 05:54, and attempt 4 continued at 05:55 from Braid's note. Cost: about 15 minutes and the judge's working context twice. Not a Braid finding, but a run started on a Saturday night meets the Sunday update window; worth scheduling runs or the window around each other. Judge checks so far (attempt 2, all at `medium`, none reading a file): lobby 13.9 min (the waits, finding 15) + continuation 1.6 min, table 6.2 min, scoreboard/help/mute 3.0 min.
Author
Owner
  1. The stall of finding 14 repeated on attempt 4, word for word: "I'll pause the clock (which holds the server's broadcast), fire start, then step the clock" → a page script awaiting the server under the pause → silent 1,200 s → watchdog redo (06:21), attempt 5. The judge has learnt from the brief that a pause holds the server's messages and reaches for it, then waits for a reply that the pause itself holds. Two things follow: the bound fix (PR #428, green) is being deployed now with a stop and resume, so the next such script is cut at 180 s; and the probe text should say plainly that nothing from the server reaches the page while the clock is paused, so a reply is never awaited under a pause (step or resume first). The second goes to the next batch with #427.

Attempts 2 to 4 of this round have cost 56 + 10 + 26 minutes; the round's four browser checks are done and in the report, so attempt 5 continues from them.

17. **The stall of finding 14 repeated on attempt 4**, word for word: "I'll pause the clock (which holds the server's broadcast), fire start, then step the clock" → a page script awaiting the server under the pause → silent 1,200 s → watchdog redo (06:21), attempt 5. The judge has learnt from the brief that a pause holds the server's messages and reaches for it, then waits for a reply that the pause itself holds. Two things follow: the bound fix (PR #428, green) is being deployed now with a stop and resume, so the next such script is cut at 180 s; and the probe text should say plainly that nothing from the server reaches the page while the clock is paused, so a reply is never awaited under a pause (step or resume first). The second goes to the next batch with #427. Attempts 2 to 4 of this round have cost 56 + 10 + 26 minutes; the round's four `browser` checks are done and in the report, so attempt 5 continues from them.
Author
Owner

Round 1 passed on the judge's first round: 34 criteria pass, the session ran end to end, no console errors, verdict pass. No fix round, so the loop closed at round 1 (run 181 needed two rounds and 156 judge minutes; run 174 two rounds and 50). The feature is committed and the summary step is running.

  1. The judge round in numbers. 204 minutes on the step across six attempts, of which about 53 were the two stalls (finding 14, 17) and two attempts were cut by the host's reboot (finding 16) and one by the stop for PR #428's deploy. The attempts that judged: 2 (56 min, 268 turns, four browser checks), 5 (58 min, 201 turns) and 6 (26 min, 98 turns, the report). Ten browser checks in all, none reading a file (#419's change holds); their lengths ran from 1.6 to 13.9 min, the long one all waits (finding 15).
  • #415 held: the clock held the dealing phase and the judge captured the deal and the trump flip as filmstrips (two osf.engine.filmstrip runs), the sweep and the scratch flash as held frames.
  • #416 did not: of the final attempt's 44 page scripts, none was the pasted motion probe; the judge wrote 36 freezes of its own around getAnimations() and the clock, and ran the pasted text probe once. The rule says "never write a freeze of your own"; it did anyway, because its own freeze handles the server round trip the probe does not. The probe should take an action that awaits a reply, or the rule should say when a freeze of one's own is right.
  • #420 held: every judge request carried medium, and the round's criteria read sound.
  • #414 (round 2's look list) cannot be measured on this run: there was no round 2.
**Round 1 passed on the judge's first round: 34 criteria pass, the session ran end to end, no console errors, verdict pass. No fix round, so the loop closed at round 1** (run 181 needed two rounds and 156 judge minutes; run 174 two rounds and 50). The feature is committed and the summary step is running. 18. **The judge round in numbers.** 204 minutes on the step across six attempts, of which about 53 were the two stalls (finding 14, 17) and two attempts were cut by the host's reboot (finding 16) and one by the stop for PR #428's deploy. The attempts that judged: 2 (56 min, 268 turns, four `browser` checks), 5 (58 min, 201 turns) and 6 (26 min, 98 turns, the report). Ten `browser` checks in all, none reading a file (#419's change holds); their lengths ran from 1.6 to 13.9 min, the long one all waits (finding 15). - **#415 held:** the clock held the dealing phase and the judge captured the deal and the trump flip as filmstrips (two `osf.engine.filmstrip` runs), the sweep and the scratch flash as held frames. - **#416 did not:** of the final attempt's 44 page scripts, none was the pasted motion probe; the judge wrote 36 freezes of its own around `getAnimations()` and the clock, and ran the pasted text probe once. The rule says "never write a freeze of your own"; it did anyway, because its own freeze handles the server round trip the probe does not. The probe should take an action that awaits a reply, or the rule should say when a freeze of one's own is right. - **#420 held:** every judge request carried `medium`, and the round's criteria read sound. - **#414 (round 2's look list) cannot be measured on this run:** there was no round 2.
Author
Owner

Run 182 finished: succeeded, its pull request merged and the change archived, 348 min wall, 345 agent minutes (run 181: 363 wall, 362 agent; run 174: 225 agent).

step run 182 run 181
explore + propose + review + revise 18 18
apply 115 (two attempts, a stop between) 148 (two attempts, a steer between)
code review 6 14
Visual check 204 (one round; 53 min of stalls, two attempts cut by the reboot, one by a deploy stop) 156 (two rounds, with a 23 min fix round)
summarize 3 2

Without the two stalls and the reboot the run would have taken about 260 agent minutes with a passing first round; with them it matched run 181's wall and did more (no fix round needed).

Scorecard, final

  • #412 steer: not exercised (no steer sent); the apply's second attempt continued from the stop as designed.
  • #413 screenshots folder: held.
  • #414 round 2's look list: not exercised (no round 2).
  • #415 clock holds socket state: held; filmstrips of the deal and the trump flip.
  • #416 pasted motion probe: did not hold; the judge wrote its own freezes (finding 18).
  • #417 gradient false positive: the text probe ran once and listed nothing; no false positive seen.
  • #418 pasted files not re-read: held (0 of 29); the paste budget is the problem (#427).
  • #419 browser checks: no file reads in 10 checks; lengths 0.4–16.8 min, three at the 60-turn cap; measurement on #419.
  • #420 judge at medium: held, with a sound first round.

New issues from this run: #424 (helper tab), #426 (review lens and the hook), #427 (paste budget), #429 (no awaiting the server under a pause). Fixed during the run: Ask's delivery (PR #423), the page-script bound on the paused clock (PR #428).

**Run 182 finished: succeeded, its pull request merged and the change archived, 348 min wall, 345 agent minutes (run 181: 363 wall, 362 agent; run 174: 225 agent).** | step | run 182 | run 181 | |---|---|---| | explore + propose + review + revise | 18 | 18 | | apply | 115 (two attempts, a stop between) | 148 (two attempts, a steer between) | | code review | 6 | 14 | | Visual check | 204 (one round; 53 min of stalls, two attempts cut by the reboot, one by a deploy stop) | 156 (two rounds, with a 23 min fix round) | | summarize | 3 | 2 | Without the two stalls and the reboot the run would have taken about 260 agent minutes with a passing first round; with them it matched run 181's wall and did more (no fix round needed). **Scorecard, final** - #412 steer: not exercised (no steer sent); the apply's second attempt continued from the stop as designed. - #413 screenshots folder: held. - #414 round 2's look list: not exercised (no round 2). - #415 clock holds socket state: held; filmstrips of the deal and the trump flip. - #416 pasted motion probe: did not hold; the judge wrote its own freezes (finding 18). - #417 gradient false positive: the text probe ran once and listed nothing; no false positive seen. - #418 pasted files not re-read: held (0 of 29); the paste budget is the problem (#427). - #419 browser checks: no file reads in 10 checks; lengths 0.4–16.8 min, three at the 60-turn cap; measurement on #419. - #420 judge at medium: held, with a sound first round. New issues from this run: #424 (helper tab), #426 (review lens and the hook), #427 (paste budget), #429 (no awaiting the server under a pause). Fixed during the run: Ask's delivery (PR #423), the page-script bound on the paused clock (PR #428).
Author
Owner

Finding 7 withdrawn. The apply's own in-browser verification (whole sessions by script, 111 page scripts, no browser subagent) was read as a cost. It is the likelier reason this run passed the judge's first round with no fix round, where runs 174 and 181 each needed a second round. The developer decides how to check its own work; no restriction follows. The one thing to carry: the apply brief's wording that says to delegate looks to browser by default is contradicted by this outcome and should be struck, not enforced. The 180 s bound on a page script that does not settle stays as the only guard.

**Finding 7 withdrawn.** The apply's own in-browser verification (whole sessions by script, 111 page scripts, no `browser` subagent) was read as a cost. It is the likelier reason this run passed the judge's first round with no fix round, where runs 174 and 181 each needed a second round. The developer decides how to check its own work; no restriction follows. The one thing to carry: the apply brief's wording that says to delegate looks to `browser` by default is contradicted by this outcome and should be struck, not enforced. The 180 s bound on a page script that does not settle stays as the only guard.
Author
Owner

Correction to finding 18 and the scorecard: no filmstrip was ever composed. The judge's one run of osf.engine.filmstrip (attempt 6, the deal's five frames) failed with Cannot find package 'playwright': the helper resolves the Playwright CLI symlink (/opt/playwright/bin/playwright -> ../lib/node_modules/playwright/cli.js) before walking up to the package and so looks in the wrong folder; the script run by hand with the package folder set composes a strip fine. The judge read four frames one by one, as the fallback told it, and never tried the trump flip's strip. So #415 held only in that the clock held the server's state; the frames were the judge's own getAnimations() freezes and the strip did not exist.

Where the findings went. Finding 7 withdrawn (above). Findings 2, 11–13: #426, #427. Findings 14, 15, 17: #419's wait cap; #429 is moot under #433. Finding 3: #430. Finding 16: the reboot was handled as designed, nothing to do. Finding 18 and the strip bug: the motion capability is reverted whole in #433, and a method is researched with a person judging alongside in #434. The wall-time analysis (docs/design/measurements/2026-11-08-runs-172-182-wall-time.md): the Visual check's list doubled because the apply left 11 state-change scenarios to it on the ground that it had looked at them itself, #431; the Visual check's browser-only framing, #432. #414 waits for a run with a fix round.

**Correction to finding 18 and the scorecard: no filmstrip was ever composed.** The judge's one run of `osf.engine.filmstrip` (attempt 6, the deal's five frames) failed with `Cannot find package 'playwright'`: the helper resolves the Playwright CLI symlink (`/opt/playwright/bin/playwright -> ../lib/node_modules/playwright/cli.js`) before walking up to the package and so looks in the wrong folder; the script run by hand with the package folder set composes a strip fine. The judge read four frames one by one, as the fallback told it, and never tried the trump flip's strip. So #415 held only in that the clock held the server's state; the frames were the judge's own `getAnimations()` freezes and the strip did not exist. **Where the findings went.** Finding 7 withdrawn (above). Findings 2, 11–13: #426, #427. Findings 14, 15, 17: #419's wait cap; #429 is moot under #433. Finding 3: #430. Finding 16: the reboot was handled as designed, nothing to do. Finding 18 and the strip bug: the motion capability is reverted whole in #433, and a method is researched with a person judging alongside in #434. The wall-time analysis (`docs/design/measurements/2026-11-08-runs-172-182-wall-time.md`): the Visual check's list doubled because the apply left 11 state-change scenarios to it on the ground that it had looked at them itself, #431; the Visual check's browser-only framing, #432. #414 waits for a run with a fix round.
Author
Owner

Closed: every finding here was filed or fixed. #426, #427, #430 (reopened and fixed again in PR 446), #431, #432 shipped in PRs 435 to 438 and 446 to 449; #433 reverted the motion capability and #434 holds the research; #419 closed on its measurement; #414 held on run 183 (its round 3 looked at 6 and carried 53). Run 183's own findings were braid#439, now closed too.

Closed: every finding here was filed or fixed. #426, #427, #430 (reopened and fixed again in PR 446), #431, #432 shipped in PRs 435 to 438 and 446 to 449; #433 reverted the motion capability and #434 holds the research; #419 closed on its measurement; #414 held on run 183 (its round 3 looked at 6 and carried 53). Run 183's own findings were braid#439, now closed too.
Sign in to join this conversation.
No labels
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set

Reference
cmoriarty/braid#425
No description provided.